TelcoFuturism - Part 7
This is the seventh blog post in a series on “TelcoFuturism” – which focuses on the intersection points between the telecom industry, areas of advanced technology, such as drones, AI, AR/VR and robotics, plus societal changes such as climate change. Disruptive Analysis looks at these adjacent technologies through the lens of what’s really happening in networking, plus the tough practicalities and complexities telcos face, rather than accepting the general hype and rose-tinted views common among some forecasters.
At first glance, it is not clear that robots have major connections to telecoms. Most people probably think first about industrial welding robots in a car factory, or perhaps a domestic vacuum cleaner or lawnmower. Beyond that, typical perceptions of robots centre on sci-fi movie characters from C3PO to the Terminator, or perhaps one of Boston Dynamics’ cool-but-creepy walking “dogs”.
Almost all of these examples have local awareness and intelligence to perform tasks offline. Roomba cleaners can work offline. Arnold Schwarzenegger’s iconic robot assassin didn’t need to connect to SkyNet with ultra-low latency wireless.
Yet robotics involves much more than these early stereotypes. Many are becoming more relevant to telecom operators. Connectivity – mostly wireless – will be central to robot’ performance and capabilities in future. We already see growing discussion of “cloud robots”. We can also expect some robots to be used within the telecoms industry itself.
Broadly speaking, robots can be divided into:
- Industrial robots, used in manufacturing or similar environments for materials-handling, welding, assembly and so on. These are mostly large, expensive and integrated into complex automation platforms. Many are fixed in specific places. Around half a million industrial robots are sold annually, with Asia accounting for a large percentage of the market. Automotive and electronics industries account for almost half of industrial robots installed.
- Professional service robots. This is a high-end segment for use primarily in medical contexts, logistics (e.g. AGVs – automatic guided vehicles - in warehouses or hospitals), security patrollers, agriculture and defence. There are also some interesting niches such as “public relations robots” (e.g. on trade-show booths), professional cleaning, or human exoskeletons.
- Domestic service robots for use in home environments, hotels and similar contexts. These are sold in much greater volumes, but for simpler tasks such as vacuum-cleaning or entertainment.
The global robotics market is worth around $50bn, and should grow strongly in future, after the current pandemic (Coronavirus) and associated global economic downturn. Various technical advances have helped drive market growth in recent years, including standardised interfaces, cloud-based control and improved ease-of-use / programming techniques. In future, we can expect machine-learning to help improve robot “training”, for instance by video-based demonstrations of tasks, and subsequent self-optimisation.
Telecom operators’ roles in robotics
Disruptive Analysis sees five broad ways in which telecom and network operators will become involved in the robot space. These include:
- Provisioning of private 5G networks in industrial, logistics or other campus settings, suitable for robot connectivity (alongside other applications)
- Creation of robot-as-a-service and “cloud robot” business models, or acting as a channel for 3rd-party RaaS
- Residential broadband and in-home connectivity suitable for robots, optionally coupled to smart-home platforms
- Enablement of “cobot” (collaborative robot) models, combining humans and robots for specific tasks.
- Robots used within the telecoms industry itself
(This excludes classes of automated machinery such as autonomous vehicles or self-flying drones, which have their own category).
Enterprise-grade networks & 5G for automation & robotics
There are close links between robotics and the success of MNOs in providing future campus / industrial private mobile networks. While some vertical and enterprise sites will deploy fully-private 4G/5G infrastructure, or work with new and specialised SPs, others will rely on traditional MNOs to provide connectivity in factories, ports, hospitals, retail sites etc.
Over time and subsequent 3GPP releases, we can expect ever-greater levels of industrial-optimised capability of 5G, to support time-sensitive deterministic networking, ultra-high throughputs and low latencies, fine-grained network slicing and so on.
That said, while wireless connections enable easier physical reconfiguration of sites, and obviously mobility for systems like AGVs, many machines and industrial locations will continue to use widespread fibre networks – either because systems are fixed in place, or because of requirements around radio interference and reliability.
These will add progressively more opportunities for control and operation of robots, either by the business itself, or as a managed service by telcos with particular vertical or horizontal specialisations or partnerships.
Various examples are being trialled. NTT DoCoMo is working with Nokia and automation specialist Omron to prove the viability of AGVs delivering products in factory environments without pre-defined paths. Singaporean operator M1 is working with a local technology university to work on 5G-connected “tactile” robots that could exploit the future ultra-low latency capabilities of 5G for security, delivery, real-time remote sensing and remote operation of robots. Many network vendors have prototype “smart factories” for their own equipment, as a showcase or testbed.
Another promising area is for robots in warehouses and logistics centres. Although again there is a lot of work by vendors of private-network solutions for stock-selection and loading robots (for example, by Ocado), there is also scope for involvement by mainstream CSPs as well – especially where service continuity is needed between a warehouse and the wide-area, when a shipment when loaded onto a truck or train.
That said, some areas seem overhyped. The oft-mentioned “surgical robots” suggested as a 5G use-case seem a long way off, not least because most hospital operating theatres are heavily shielded and deep in-building. It is perhaps easier to imagine an emergency paramedic or military field-hospital setting for such devices.
Cloud robots & Robots-as-a-Service
An interesting recent development in robotics is cloud-based platforms for robotics, which can offload some of the compute, storage and AI functions from individual bots to a shared platform, running in a remote datacentre or edge cloud.
This allows both a reduction of costs in the robots themselves, as well as more-efficient use and analysis of shared data to optimise their activities and behaviour. They can centralise functions such as task-planning, mapping and visual identification, models of local environments and so on. This also allows coordination of multiple robots or “swarms”.
Some telecom operators are already looking at cloud-based robots, and the associated RaaS (robot as a service) business models, which do not require individual enterprises to acquire and own robots directly. This makes automated systems more accessible to smaller and mid-sized businesses with less upfront capital outlay, and allows for easier trials and experimentation.
- BICS is partnering with German Bionic, which provides connected human exoskeletons with a robot-as-a-service business model
- China Mobile is working with CloudMinds Technology, which provides remote-controlled robots for various tasks, most notably disinfection and disease-control functions during the COVID-19 outbreak in Wuhan and Shanghai hospitals.
It seems likely that RaaS-type services will fit best with telcos’ existing vertical and horizontal specialisms – perhaps linked to domains such a smart cities (e.g. for cleaning), retail and business property (security patrols) and healthcare.
Domestic robots, broadband & smart home services
Most current domestic robots are simple, Wi-Fi connected products used for vacuum cleaning, lawn mowing and basic entertainment. Some can work offline, just connecting periodically for software updates, or some forms of non-real time data sharing.
This may change significantly in future, especially as changing demographics bring the need for more personal-care, elder-care and medical robots into the home. We may also see a rise in consumer-grade cleaning / disinfecting robots in the wake of the COVID-19 pandemic. Other possibilities include new forms of robot “pet” and perhaps others for food-preparation, maintenance of vehicles, DIY tasks and other functions.
This evolution path could make consumer robotics more telco-friendly. Options include:
- Delivery of cloud-robot and managed-service functions for residential subscribers, as part of smart home packages.
- Ability to offer QoS and managed security, through fixed/Wi-Fi or cellular connections, suitable for medical or care robots for vulnerable individuals.
- “Domestic” / personal robot use will increasingly extend beyond the home, requiring wide-area wireless connectivity.
- Shared community- or near-enterprise grade service robots in larger residential settings, such as retirement communities, villages or large apartment blocks. These require network coverage and some form of service/billing model.
Human / Robot collaboration & working
An early-stage trend is the combination of humans and robots for collaborative applications. This can vary from shared workspaces (with appropriate safety controls), to direct co-working on the same task. This could be either inside a factory, or in the field – for example on construction sites, or in emergency settings.
There are also various domestic use-cases, especially where a robot is actually performing tasks on a human, such as personal care or health-related applications. That said, “cobots” are still just a tiny fraction of the overall robot marketplace today, although this is expected to grow quickly in coming years.
Telcos have a number of roles to play here. Perhaps the most important involves positioning technologies and services. 5G in particular enables much more accurate, absolute location and proximity sensing than previous technologies, which may be used to enhance safety. (In 3GPP R17 there should be cm-level accuracy).
There may also be ways to tie together humans and robots from a security or compliance perspective – for example, using device identities or SIMs to “pair” the two workers, or allow “sign-in” for tasks.
We will also see remote collaboration – for example a paramedic robot being partially autonomous, and partly under the control of a medical professional at a distance. A similar model could apply in dangerous or confined spaces. Clearly, this would require high bandwidth and low latencies to work effectively – and perhaps, also various edge-compute capabilities for control.
Robots in telecoms
The telecoms industry is often good at projecting how new technology trends drive new services or revenue streams. It is less focused on how those trends could play internally.
There is comparatively little discussion of robots being used in the telecoms sector itself. While physical operations are not always as easy to automate because of technical complexity, the following areas are important now, or potentially in future:
- Widespread use of robots in the manufacture and testing of smartphones and other wireless devices, as well as some classes of network equipment hardware.
- Routine field-service tasks such as inspection of equipment (and ducting / poles), creating or maintaining connections in street cabinets. Specialised units can inspect or repair subsea cables.
- Tasks within telcos’ datacentres, warehouses or other facilities, such as security monitoring, AGVs and basic maintenance.
- Repetitive tasks in lab environments, such as setting up and reconfiguring test systems.
- Network monitoring or testing tasks (such as urban or in-building “drive” tests) conducted with robots.
- In the medium term, it may become possible for robots to install radio units on towers, dig trenches (perhaps with remote human collaboration) and deploy other infrastructure.